There are millions of posts to Facebook everyday and there is no way for Facebook to police all of them, so they have turned to AI to help scan for inappropriate behavior on their platform. Algorithms have been sucessful at detecting nudity, but they have not been nearly as successful in detecting bullying and harrassment-86% of the incidents were reported by users. Facebook acknowledged they need to get better with identifying bullying and that it is a priority for them, yet incidents such as those in the article that clearly violate Facebook's policies and are consistent with their guidelines for deactivating accounts continue to occur. In one of the incidents described in the article, the harassee reported the incident to Facebook and the harasser's only punishment was to be prohibited from using Messenger for 30 days. It wasn't until CNN investigated this (and other) incidents and questioned why the harrassers weren't removed from the platform, that FB then closed the account.
Yes, algorithms are needed to help 'police' the platform, but the problem then goes to what happens when wrongdoing is detected? Why aren't obvious, reported claims of harassment and bullying beind handled according to FB's policies? Zuckerberg spoke last year about making this a priority. Perhaps it's time to have the programmers focus more time on the algorithms and beef up the team that then reviews and handles the incidents.
https://apple.news/AJ2N8PkGURBGV77zau2R-GQ
Yes, algorithms are needed to help 'police' the platform, but the problem then goes to what happens when wrongdoing is detected? Why aren't obvious, reported claims of harassment and bullying beind handled according to FB's policies? Zuckerberg spoke last year about making this a priority. Perhaps it's time to have the programmers focus more time on the algorithms and beef up the team that then reviews and handles the incidents.
https://apple.news/AJ2N8PkGURBGV77zau2R-GQ
No comments:
Post a Comment